latigid on

I am thinking about a control circuit for CVs in any one of the available AOUT modules. The concept involves a -10 to 0 V source derived from the DAC and an inverting amplifier. The second op amp stage can do one of three things: 1. Inverting buffer (voltage level = 0 to 10 V) 2. Inverting amplifier (voltage level = 0 to 5 V) 3. Inverting mixer (voltage level = -5 to 5 V) This would be accomplished by an ON-OFF-ON DPDT toggle switch. Case 1 allows the voltage to pass without any further change (middle position). The second switches in a resistor in parallel with the feedback element. Using a cermet trimmer, one can get an accurate gain setting of 0.5 for unipolar 0-5 V (or any other value up to 10 V). The third adds a trimmed offset of 5 V for bipolar operation. Should it work? Thoughts and criticisms are welcome. :)

Also a bit too big for my use also, sorry!

Possibly interested in one, depending on pics!

Really like this one Peter, keep it up!

Nice work!

Re: connectors, I guess it depends on which core you'll use! For the TTSH I have a gate booster pcb that also generates a trigger pulse kindly designed and routed/fabbed by people on MW. But I was originally thinking of similar oddball synths like the 2600 when I was playing with CV/gate breakout boards a year ago. It's still nice to have higher voltage available for non-Doepfer gear. My comment on voltage regulation was more applicable to systems with bussed power. If you have a regulated bipolar supply dedicated to this AOUT module, no problem. I hear that taking your VCO pitch reference directly from a power rail is not good design.

You should probably discuss this more in the HUI section of the forum, this part is for new design concepts.

Probably a good idea to start going SMT for small-scale applications! Can I ask why you have decided on the 525 as a DAC? Better linearity than the TLV5630? really progressed this year, I moved and now I'm occupied with Eurorack projects. But, I still plan on using a SEQ V4 and/or a with analogue outs. Maybe I was too optimistic about getting the power from the modular though and I'm toying with the idea of a larger (3U x 126 HP) SEQ case with outputs mounted topside. The idea of trimming the gate down from 12 V is a good one, it might also be a good idea to provide a regulated voltage for the offset control. as it makes bipolar trimming very difficult or impossible. Perhaps a 1:1 header pinning from the core could be useful if there's space? And you might also want to keep the digital and analogue sections of the PCB separate, which is harder than it sounds!

C'est trop cher!

Can you upload a picture with the ground plane poured? Some connections look a bit funny.

Did you correctly change the sequencer output to AOUT1-8? The default will be USB, MIDI, IIC etc. Failing that you could try the TEST AOUT PIN program from the Core, it should be somewhere on ucapps.de Bonne chance, Andy

Check your J5 and resistor network orientations too!

Nice, have improvements been made with regard to LED saturation?

sneakthief, I see some issues with one standard of input: If a bipolar CV (like an LFO) is plugged in, the negative portion of the waveform will be clipped to ~0 V by the Schottky diode. Also for nominal signals of say 0-5 V you lose half the ADC resolution as the full scale input is 10 V. But the option should be there to choose your own pre-ADC treatment. I will get around to a board layout at some point, just a bit busy at the moment.

Wow, looks awesome! So, now we have a gate to MIDI converter -- any thoughts on CV-MIDI conversion at 1 V/octave? With the AIN scaling/protection circuits it seems to be difficult to decide on one standard, so I propose on a final panel a mixture of 0-5, 0-10 and -5 to +5 inputs, hopefully that's workable with the correct offsets. Alternatively each input could have an attenuator: http://www.falstad.com/circuit/#%24+1+4.9999999999999996E-6+47.65948060424672+58+15.0+60%0Ar+640+352+704+352+0+20000.0%0Az+912+352+912+304+1+0.805904783+3.3%0Az+912+400+912+352+1+0.805904783+3.3%0Ag+912+400+912+432+0%0AR+912+304+912+272+0+0+40.0+3.3+0.0+0.0+0.5%0AR+400+336+368+336+0+4+50.0+10.0+10.0+0.0+0.5%0Ap+912+352+960+352+1%0Aa+528+352+624+352+0+15.0+-15.0+1000000.0%0Ag+528+368+528+400+0%0Ar+528+336+480+336+0+20000.0%0Ar+464+336+400+336+0+20000.0%0Ar+496+432+496+496+0+100.0%0Ar+544+304+592+304+0+20000.0%0Ar+592+304+640+304+0+20000.0%0Aw+640+352+624+352+0%0Aw+528+336+528+304+0%0Aw+528+304+544+304+0%0Ag+400+432+400+448+0%0Aa+736+352+816+352+0+15.0+-15.0+1000000.0%0Aw+736+368+736+400+0%0Aw+736+352+736+336+0%0Ag+736+400+736+432+0%0Aw+736+352+704+352+0%0Aw+736+336+736+288+0%0Ar+736+288+832+288+0+20000.0%0Aw+816+352+816+320+0%0Aw+816+320+832+320+0%0Aw+832+320+832+288+0%0Aw+640+352+640+304+0%0Ar+816+352+912+352+0+100.0%0A174+400+368+448+432+0+10000.0+0.9455000000000001+Resistance%0Aw+480+336+464+336+0%0Aw+464+400+464+336+0%0Aw+464+400+448+400+0%0Ao+5+64+0+39+20.0+0.0015625+0+-1%0Ao+6+64+0+38+5.0+9.765625E-5+0+-1%0A'>Falstad sim, never a bad thing with modular patches.

You might've fried your LCD, that's easy to do. You describe shorting the backlight, well if you've applied a high current across any of the data lines that could knock out some of the circuitry. Do you have the option of using a second backlight for a test? If so, be sure you don't have any shorts that might repeat the process and outcome.

With the discussion on the CV V2 topic I'm thinking about what this means for the SEQ application. I would like to have both in a modular set up: MBSEQ for, well, sequences, and MBCV for interesting envelopes, LFOs, bussed knob controllers, scopes etc. There's a bit of overlap: the SEQ sends out 8 channels of CV/Gate and DIN SYNC while the MBCV can additionally output more clock signals and receive analogue inputs and a clock as control sources. The use of balanced line drivers will be useful for both boxes for AOUT, but how about the digital signals? The hardware config file for the SEQ lists SR outputs for gates, so is there a possibility of avoiding J5A/B etc. and just sending the balanced DOUT SRIO chain instead? And what about LED indication for the gates? Could a different SR output a 5 V signal to turn a gate on and off? Reading the config file maybe this is not possible, what is the explanation? The parts count would be a bit lower than using common emitter LED drivers.

Try here: http://www.muffwiggler.com/forum/viewtopic.php?t=102469&postdays=0&postorder=asc&start=0

Hmm, I had a try at putting 32 sockets on a 4.05" board (basically the max. height for a Eurorack PCB): 8 gate 8 CV 8 AIN 7 Clock 1 Sync in and it's pretty tight even at 2.2" (for a 12 HP panel), especially with another multipin connector and mounting holes. I suggest a DB-25 (or whatever connector) is mounted on a PCB with all of the interfacing circuitry (line receivers, analogue input protection circuits, pass-through for AOUT and DOUT). All of the other sockets could be panel mounted and wired to MTA 100-style connectors or similar. I know the trend is towards super shallow modules with no panel wiring but this would give a better return on investment for PCB space. Also it is customisable and more flexible for other modular formats (5U, Serge etc.) or even standalone MBCV enclosures.

Quick and naive question: are the serial clocks independent; could they be shared too?

I was considering a dual use board, one for MBseq and one for MBCV. This is for us old farts who are still on STM32F1 cores! So we could get away with a 15 pin connector for the digital serial signals and a 9 pin for AINs.